|
Recently, the study of the mathematical model has generated the development of a wide range of analysis of the randomness complex systems involving multiple field, such as the more common method of complex biological, environmental, economic, financial, and material processing engineering. system build differential equation model, combined with computer simulation techniques to explore the dynamic evolution of the system, and thus to predict, organize and control system. inverse problem using the observational data of differential equations, differential equations parameter estimation, International The sector research focus on probability and statistics. many scholars in research Differential parameters estimated are prey - predator model simulated objects, in order to analyze the validity of the estimation method. therefore, it is necessary to Predator - Prey Model parameter estimates for more in-depth and detailed discussion of this paper, the prey - predator model parameter estimation, a new two-stage estimation method of the first stage, based on the finite difference principle, the use of observational data for solving numerical differentiation, the second stage of the prey - predator model parameter estimation problems into a linear least squares problem. two-stage method greatly reduces the amount of computation, and improve computing speed. then a more detailed theory of the two-stage method research, given the statistical structure of the differential equation parameter estimation in the general sense, and to discuss the large sample properties of the two-stage method. Finally, the numerical simulation of the two species prey - predator model parameter estimation, the results show that, this paper presents The two-stage method is effective. consists of five chapters, the first chapter briefly prior knowledge of the problem of the historical background, the article relates to; Chapter II first given theoretically general differential equation parameters estimated statistical structure, and discuss in detail the nonlinear least squares estimation, parameter estimation method based on finite difference numerical differentiation stage; Chapter consistency and asymptotic normality of two-stage parameter estimators Large sample properties; Chapter Two Species Predator - Prey Model parameters estimates numerical simulation, detailed analysis of three common finite difference numerical differentiation, parameter estimation, and nonlinear least squares estimation and literature in the results were compared, the results show that the proposed method is feasible; Chapter made a summary of the full text and research prospects.
|